Subagent-driven development

Executes a plan through subagents: a fresh implementer and reviewer for every task, checking spec compliance and code quality, up to five fix rounds and a final review of the whole branch. Use for a plan of independent tasks when each one needs a review.

CodingPack: Superpowers339 linessubagent-driven-development
Author
Source
obra/superpowers/skills/subagent-driven-development @ 22f0b46
License
MIT · Copyright (c) 2025 Jesse Vincent

About this skill

What it does

Subagent-driven development turns the agent into a controller for a plan from writing-plans. For each task it dispatches a fresh implementer subagent, then a fresh reviewer that returns two verdicts, spec compliance and code quality. A reviewer on the most capable model checks the whole branch at the end. Subagents never see the session history: they get a brief file and the constraints, report into files so the controller's context stays small, and may not spawn subagents of their own.

A failed review opens a fix loop of at most five rounds: the original implementer in rounds 1–3, a fresh one on a stronger model in rounds 4–5, each round followed by a scoped re-review. What is still open after that is parked with a ruling in the progress.md ledger; the same ledger lets the controller resume after compaction.

The folder holds three prompt templates (implementer-prompt.md, task-reviewer-prompt.md, re-review-prompt.md), seven references/ files read at set moments, and four bash scripts: sdd-workspace (the plan's git-ignored folder under .superpowers/sdd/), sdd-repo (which repository to work in), task-brief (one task cut out of the plan into a file) and review-package (a commit range's log and diff in one file for the reviewer).

When to use it, and when not

Use it for a plan with mostly independent tasks when your tool can dispatch subagents and you want a review gate on every task.

Skip it for tightly coupled tasks (the skill sends those to manual work or back to brainstorming), and use executing-plans when per-task review is not worth its cost.

How to run it

The install command below puts the whole folder in place. The agent picks the skill up by its description when a plan is ready; to call it yourself:

  • Claude Code: /subagent-driven-development, or ask for the skill by name.
  • Codex: $subagent-driven-development.
  • Cursor: type / in the Agent chat and pick subagent-driven-development.

Example: /subagent-driven-development Run docs/superpowers/plans/2026-11-20-csv-export.md, cheap model for mechanical tasks.

The final review uses code-reviewer.md from requesting-code-review, so install it in the same skills folder. The work starts with using-git-worktrees and ends with finishing-a-development-branch. executing-plans borrows this skill's scripts and needs it installed. On Windows the scripts need Git Bash.

How it differs from similar skills

executing-plans runs the same plan in one context with a single review at the end: cheaper, no per-task gate. dispatching-parallel-agents sends agents at independent problems at the same time; here implementers go strictly in turn, because parallel ones conflict. Implement a spec does run implementers in parallel, each in its own worktree, following the ticket graph, and reviews once at the end. PR review and Self-review before PR from our catalog check a finished diff once; this skill reviews every task and re-reviews every fix.

Install

Run one command — it saves SKILL.md and the files it uses into your agent's skills folder. Claude Code and Codex pick up a new skill without a restart; if it does not show up, restart the agent.

Claude Code — macOS, Linux
curl -fsSL --create-dirs -o ~/.claude/skills/subagent-driven-development/SKILL.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/SKILL.md -o ~/.claude/skills/subagent-driven-development/LICENSE https://anymodel.org/en/skills/superpowers/subagent-driven-development/LICENSE -o ~/.claude/skills/subagent-driven-development/implementer-prompt.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/implementer-prompt.md -o ~/.claude/skills/subagent-driven-development/re-review-prompt.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/re-review-prompt.md -o ~/.claude/skills/subagent-driven-development/references/batching.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/batching.md -o ~/.claude/skills/subagent-driven-development/references/cannot-verify.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/cannot-verify.md -o ~/.claude/skills/subagent-driven-development/references/final-review.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/final-review.md -o ~/.claude/skills/subagent-driven-development/references/fix-loop.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/fix-loop.md -o ~/.claude/skills/subagent-driven-development/references/handling-reports.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/handling-reports.md -o ~/.claude/skills/subagent-driven-development/references/preflight.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/preflight.md -o ~/.claude/skills/subagent-driven-development/references/resuming.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/resuming.md -o ~/.claude/skills/subagent-driven-development/scripts/review-package https://anymodel.org/en/skills/superpowers/subagent-driven-development/scripts/review-package -o ~/.claude/skills/subagent-driven-development/scripts/sdd-repo https://anymodel.org/en/skills/superpowers/subagent-driven-development/scripts/sdd-repo -o ~/.claude/skills/subagent-driven-development/scripts/sdd-workspace https://anymodel.org/en/skills/superpowers/subagent-driven-development/scripts/sdd-workspace -o ~/.claude/skills/subagent-driven-development/scripts/task-brief https://anymodel.org/en/skills/superpowers/subagent-driven-development/scripts/task-brief -o ~/.claude/skills/subagent-driven-development/task-reviewer-prompt.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/task-reviewer-prompt.md && chmod +x ~/.claude/skills/subagent-driven-development/scripts/review-package ~/.claude/skills/subagent-driven-development/scripts/sdd-repo ~/.claude/skills/subagent-driven-development/scripts/sdd-workspace ~/.claude/skills/subagent-driven-development/scripts/task-brief
Codex, Cursor — macOS, Linux
curl -fsSL --create-dirs -o ~/.agents/skills/subagent-driven-development/SKILL.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/SKILL.md -o ~/.agents/skills/subagent-driven-development/LICENSE https://anymodel.org/en/skills/superpowers/subagent-driven-development/LICENSE -o ~/.agents/skills/subagent-driven-development/implementer-prompt.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/implementer-prompt.md -o ~/.agents/skills/subagent-driven-development/re-review-prompt.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/re-review-prompt.md -o ~/.agents/skills/subagent-driven-development/references/batching.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/batching.md -o ~/.agents/skills/subagent-driven-development/references/cannot-verify.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/cannot-verify.md -o ~/.agents/skills/subagent-driven-development/references/final-review.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/final-review.md -o ~/.agents/skills/subagent-driven-development/references/fix-loop.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/fix-loop.md -o ~/.agents/skills/subagent-driven-development/references/handling-reports.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/handling-reports.md -o ~/.agents/skills/subagent-driven-development/references/preflight.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/preflight.md -o ~/.agents/skills/subagent-driven-development/references/resuming.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/references/resuming.md -o ~/.agents/skills/subagent-driven-development/scripts/review-package https://anymodel.org/en/skills/superpowers/subagent-driven-development/scripts/review-package -o ~/.agents/skills/subagent-driven-development/scripts/sdd-repo https://anymodel.org/en/skills/superpowers/subagent-driven-development/scripts/sdd-repo -o ~/.agents/skills/subagent-driven-development/scripts/sdd-workspace https://anymodel.org/en/skills/superpowers/subagent-driven-development/scripts/sdd-workspace -o ~/.agents/skills/subagent-driven-development/scripts/task-brief https://anymodel.org/en/skills/superpowers/subagent-driven-development/scripts/task-brief -o ~/.agents/skills/subagent-driven-development/task-reviewer-prompt.md https://anymodel.org/en/skills/superpowers/subagent-driven-development/task-reviewer-prompt.md && chmod +x ~/.agents/skills/subagent-driven-development/scripts/review-package ~/.agents/skills/subagent-driven-development/scripts/sdd-repo ~/.agents/skills/subagent-driven-development/scripts/sdd-workspace ~/.agents/skills/subagent-driven-development/scripts/task-brief
Claude Code — Windows PowerShell
foreach ($f in 'SKILL.md','LICENSE','implementer-prompt.md','re-review-prompt.md','references/batching.md','references/cannot-verify.md','references/final-review.md','references/fix-loop.md','references/handling-reports.md','references/preflight.md','references/resuming.md','scripts/review-package','scripts/sdd-repo','scripts/sdd-workspace','scripts/task-brief','task-reviewer-prompt.md') { $p = Join-Path "$HOME\.claude\skills\subagent-driven-development" $f; New-Item -ItemType Directory -Force (Split-Path $p) | Out-Null; Invoke-WebRequest "https://anymodel.org/en/skills/superpowers/subagent-driven-development/$f" -UseBasicParsing -OutFile $p }
Codex, Cursor — Windows PowerShell
foreach ($f in 'SKILL.md','LICENSE','implementer-prompt.md','re-review-prompt.md','references/batching.md','references/cannot-verify.md','references/final-review.md','references/fix-loop.md','references/handling-reports.md','references/preflight.md','references/resuming.md','scripts/review-package','scripts/sdd-repo','scripts/sdd-workspace','scripts/task-brief','task-reviewer-prompt.md') { $p = Join-Path "$HOME\.agents\skills\subagent-driven-development" $f; New-Item -ItemType Directory -Force (Split-Path $p) | Out-Null; Invoke-WebRequest "https://anymodel.org/en/skills/superpowers/subagent-driven-development/$f" -UseBasicParsing -OutFile $p }

Any other agent: save the file into its skills folder, or give it the link to the file: https://anymodel.org/en/skills/superpowers/subagent-driven-development/SKILL.md

SKILL.md

Raw

Subagent-Driven Development

Execute plan by dispatching a fresh implementer subagent per task, a task review (spec compliance + code quality) after each, and a broad whole-branch review at the end.

Why subagents: You delegate tasks to specialized agents with isolated context. By precisely crafting their instructions and context, you ensure they stay focused and succeed at their task. They should never inherit your session's context or history — you construct exactly what they need. This also preserves your own context for coordination work.

Core principle: Fresh subagent per task + task review (spec + quality) + broad final review = high quality, fast iteration

Narration: between tool calls, narrate at most one short line — the ledger and the tool results carry the record.

Continuous execution: Do not pause to check in with your human partner between tasks. Execute all tasks from the plan without stopping. The only reasons to stop are the four named below, or all tasks complete. "Should I continue?" prompts and progress summaries waste their time — they asked you to execute the plan, so execute it.

Rulings, not stalls. A running plan does not wait on a human. Conflicts, ambiguities, plan defects, a cap you would have asked to exceed — decide them. The spec is the binding authority, the plan is its argument, and your judgment settles what neither answers. Record every decision in the ledger as Ruling: <what you decided> — <why> — <what it costs if wrong>, and keep going. A wrong ruling costs rework your human partner can see and undo; a session parked on a question costs their whole day and buys nothing.

Four things stop you, and only these: an irreversible or destructive operation; a security-sensitive action; a side effect outside this worktree that norms say you ask about first (a merge, a push to a shared branch, a publish); and a plan so broken that every path forward is a guess. For those, stop and ask.

When to Use

digraph when_to_use {
    "Have implementation plan?" [shape=diamond];
    "Tasks mostly independent?" [shape=diamond];
    "Partner chose inline, or no subagent tool?" [shape=diamond];
    "subagent-driven-development" [shape=box];
    "executing-plans" [shape=box];
    "Manual execution or brainstorm first" [shape=box];

    "Have implementation plan?" -> "Tasks mostly independent?" [label="yes"];
    "Have implementation plan?" -> "Manual execution or brainstorm first" [label="no"];
    "Tasks mostly independent?" -> "Partner chose inline, or no subagent tool?" [label="yes"];
    "Tasks mostly independent?" -> "Manual execution or brainstorm first" [label="no - tightly coupled"];
    "Partner chose inline, or no subagent tool?" -> "executing-plans" [label="yes"];
    "Partner chose inline, or no subagent tool?" -> "subagent-driven-development" [label="no"];
}

vs. Executing Plans (inline):

  • Fresh subagent per task (no context pollution) instead of one context doing every task
  • Review after each task (spec compliance + code quality) instead of only at the end
  • Costs a fresh context per task and per review; inline costs one context plus one final reviewer
  • Both run in this session, share the same plan workspace and ledger, and never pause between tasks

The Process

digraph process {
    rankdir=TB;

    subgraph cluster_per_task {
        label="Per Task";
        "Dispatch implementer subagent (./implementer-prompt.md)" [shape=box];
        "Implementer asks questions?" [shape=diamond];
        "Answer questions, provide context" [shape=box];
        "Implementer implements, tests, commits, self-reviews" [shape=box];
        "Generate review package, dispatch task reviewer (./task-reviewer-prompt.md)" [shape=box];
        "Spec ✅ and quality approved?" [shape=diamond];
        "Fix loop: read references/fix-loop.md before every round" [shape=box];
        "Append completion to ledger, mark todo complete" [shape=box];
    }

    "Setup: worktree, ledger check, read plan, pre-flight review" [shape=box];
    "More tasks remain?" [shape=diamond];
    "Final review and finish: read references/final-review.md" [shape=box];
    "Final review clean: delete this plan's workspace" [shape=box];
    "Use superpowers:finishing-a-development-branch" [shape=box style=filled fillcolor=lightgreen];

    "Setup: worktree, ledger check, read plan, pre-flight review" -> "Dispatch implementer subagent (./implementer-prompt.md)";
    "Dispatch implementer subagent (./implementer-prompt.md)" -> "Implementer asks questions?";
    "Implementer asks questions?" -> "Answer questions, provide context" [label="yes"];
    "Answer questions, provide context" -> "Implementer implements, tests, commits, self-reviews";
    "Implementer asks questions?" -> "Implementer implements, tests, commits, self-reviews" [label="no"];
    "Implementer implements, tests, commits, self-reviews" -> "Generate review package, dispatch task reviewer (./task-reviewer-prompt.md)";
    "Generate review package, dispatch task reviewer (./task-reviewer-prompt.md)" -> "Spec ✅ and quality approved?";
    "Spec ✅ and quality approved?" -> "Append completion to ledger, mark todo complete" [label="yes"];
    "Spec ✅ and quality approved?" -> "Fix loop: read references/fix-loop.md before every round" [label="no"];
    "Fix loop: read references/fix-loop.md before every round" -> "Append completion to ledger, mark todo complete" [label="addressed, or parked at the cap"];
    "Append completion to ledger, mark todo complete" -> "More tasks remain?";
    "More tasks remain?" -> "Dispatch implementer subagent (./implementer-prompt.md)" [label="yes"];
    "More tasks remain?" -> "Final review and finish: read references/final-review.md" [label="no"];
    "Final review and finish: read references/final-review.md" -> "Final review clean: delete this plan's workspace";
    "Final review clean: delete this plan's workspace" -> "Use superpowers:finishing-a-development-branch";
}

Setup

Ensure the work happens in an isolated workspace: use superpowers:using-git-worktrees to create one or verify the existing one. Never start implementation on a main/master branch without your human partner's explicit consent.

Conversation memory does not survive compaction. In real sessions, controllers that lost their place have re-dispatched entire completed task sequences — the single most expensive failure observed. Track progress in a ledger file, not only in todos.

  • Each plan owns a workspace: at skill start, run this skill's bash scripts/sdd-workspace PLAN_FILE by its absolute path, from the working tree you're executing in — it prints the plan's git-ignored directory (under <repo-root>/.superpowers/sdd/), home to every artifact for THIS plan: ledger, briefs, reports, review packages. Another plan's directory is never yours to read or write.
  • Check for this plan's ledger at <workspace>/progress.md. If one exists, read references/resuming.md before anything else: it says whose ledger it is and where to resume.
  • Create the ledger with its identity as the first line: # SDD ledger — plan: <plan file path>.
  • The ledger is your recovery map: the commits it names exist in git even when your context no longer remembers creating them. After compaction, trust the ledger and git log over your own recollection.
  • git clean -fdx will destroy the workspace (it's git-ignored scratch); if that happens, recover from git log.

Read the plan once, note its context and Global Constraints, and create a todo per task. If the plan names a Spec, read that too: the spec is the authority the plan argues from, and conflicts inside the plan resolve against it. A plan with no reachable spec gets a ledger note saying so — rulings made without one are provisional.

Before dispatching Task 1, read references/preflight.md and run the conflict scan it describes. Its table and rulings go in the ledger.

Model Selection

Use the least powerful model that can handle each role to conserve cost and increase speed.

Mechanical implementation tasks (isolated functions, clear specs, 1-2 files): use a fast, cheap model. Most implementation tasks are mechanical when the plan is well-specified.

Integration and judgment tasks (multi-file coordination, pattern matching, debugging): use a standard model.

Architecture and design tasks: use the most capable available model. The final whole-branch review is one of these — dispatch it on the most capable available model, not the session default.

Review tasks: choose the model with the same judgment, scaled to the diff's size, complexity, and risk. A small mechanical diff does not need the most capable model; a subtle concurrency change does. Scoped re-reviews of small fix diffs take a cheap-to-mid tier.

Fix-loop escalation (rounds 4-5): use a model at least one tier above the implementer that got stuck.

Always specify the model explicitly when dispatching a subagent. An omitted model inherits your session's model — often the most capable and most expensive — which silently defeats this section.

Turn count beats token price. Wall-clock and context cost scale with how many turns a subagent takes, and the cheapest models routinely take 2-3× the turns on multi-step work — costing more overall. Use a mid-tier model as the floor for reviewers and for implementers working from prose descriptions. When the task's plan text contains the complete code to write, the implementation is transcription plus testing: use the cheapest tier for that implementer. Single-file mechanical fixes also take the cheapest tier.

Task complexity signals (implementation tasks):

  • Touches 1-2 files with a complete spec → cheap model
  • Touches multiple files with integration concerns → standard model
  • Requires design judgment or broad codebase understanding → most capable model

The Task Loop

Small same-shape work: when the plan lists several small edits of the same kind across files, read references/batching.md before dispatching them.

Everything you paste into a dispatch prompt — and everything a subagent prints back — stays resident in your context for the rest of the session and is re-read on every later turn. Hand artifacts over as files.

Waiting on dispatched subagents: never poll a wait interface with short timeouts, and never sit in one silent, open-ended wait either. While you have local work — ledger updates, packaging the next review, reading reports — keep working; child results arrive on their own. When you are genuinely idle, wait in bounded stretches (five to ten minutes, where your platform allows), and between stretches post one line of status and reconcile your live children: list them, and chase any that finished without reporting. A bounded stretch keeps nearly all of a long wait's efficiency while guaranteeing a stuck or lost child is noticed within minutes, not at the end of the session.

1. Dispatch the implementer

Record BASE (git rev-parse HEAD) before dispatching — the review package and fix-round diffs need it.

  • Task brief: before dispatching an implementer, run this skill's bash scripts/task-brief PLAN_FILE N by its absolute path, from the working tree you're executing in — it extracts the task's full text to a uniquely named file and prints the path. Compose the dispatch so the brief stays the single source of requirements. Your dispatch should contain: (1) one line on where this task fits in the project; (2) the brief path, introduced as "read this first — it is your requirements, with the exact values to use verbatim"; (3) interfaces and decisions from earlier tasks that the brief cannot know; (4) your resolution of any ambiguity you noticed in the brief; (5) the report-file path and report contract. Exact values (numbers, magic strings, signatures, test cases) appear only in the brief. Never make a subagent read the whole plan file.
  • Report file: name the implementer's report file after the brief (brief …/task-N-brief.md → report …/task-N-report.md) and put it in the dispatch prompt. The implementer writes the full report there and returns only status, commits, a one-line test summary, and concerns.
  • A dispatch prompt describes one task, not the session's history. Do not paste accumulated prior-task summaries ("state after Tasks 1-3") into later dispatches — a real session's dispatch hit 42k chars of which 99% was pasted history. A fresh subagent needs its task, the interfaces it touches, and the global constraints. Nothing else.
  • The dispatch carries the no-subagents contract (it is in the implementer template): the implementer never dispatches subagents — not helpers, and never a reviewer. Review arrives from you, after the report. In real sessions, every reviewer a worker spawned duplicated the task review the controller dispatched anyway — a full extra review seat per task.
  • If an earlier task parked a finding in the area this task touches, carry a pointer to that ledger entry in the dispatch.
  • Record the implementer's agent identity from the dispatch result — fix-loop rounds 1-3 resume this agent.
  • Never dispatch multiple implementation subagents in parallel (conflicts).

Template: implementer-prompt.md

2. Handle the report

Implementer subagents report one of four statuses. Handle each appropriately:

DONE: Generate the review package (bash scripts/review-package PLAN_FILE BASE HEAD by its absolute path, from the working tree you're executing in — it prints the unique file path it wrote; BASE is the commit you recorded before dispatching the implementer — never HEAD~1, which silently drops all but the last commit of a multi-commit task), then dispatch the task reviewer with the printed path.

Any other status (DONE_WITH_CONCERNS, NEEDS_CONTEXT, BLOCKED): read references/handling-reports.md and follow it, every time you get one.

If the implementer asks questions — before starting or mid-task — answer clearly and completely, provide additional context if needed, and don't rush it into implementation.

3. Review the task

Per-task reviews are task-scoped gates. The broad review happens once, at the final whole-branch review. Never skip the task review, and never accept a report missing either verdict — spec compliance AND task quality are both required. Implementer self-review never replaces the task review; both are needed.

  • Hand the reviewer its diff as a file: run this skill's bash scripts/review-package PLAN_FILE BASE HEAD by its absolute path, from the working tree you're executing in, and pass the reviewer the file path it prints (or, without bash: git log --oneline, git diff --stat, and git diff -U10 for the range, redirected to one uniquely named file). The output never enters your own context, and the reviewer sees the commit list, stat summary, and full diff with context in one Read call. Use the BASE you recorded before dispatching the implementer — never HEAD~1, which silently truncates multi-commit tasks. Never dispatch a task reviewer without a diff file.
  • Reviewer inputs: the task reviewer gets three paths — the same brief file, the report file, and the review package — plus the global constraints that bind the task.
  • The global-constraints block you hand the reviewer is its attention lens. Copy the binding requirements verbatim from the plan's Global Constraints section or the spec: exact values, exact formats, and the stated relationships between components ("same layout as X", "matches Y"). The reviewer's template already carries the process rules (YAGNI, test hygiene, review method) — the constraints block is for what THIS project's spec demands.
  • Do not add open-ended directives like "check all uses" or "run race tests if useful" without a concrete, task-specific reason
  • Do not ask a reviewer to re-run tests the implementer already ran on the same code — the implementer's report carries the test evidence
  • Do not pre-judge findings for the reviewer — never instruct a reviewer to ignore or not flag a specific issue. If you believe a finding would be a false positive, let the reviewer raise it and adjudicate it in the review loop. If the prompt you are writing contains "do not flag," "don't treat X as a defect," "at most Minor," or "the plan chose" — stop: you are pre-judging, usually to spare yourself a review loop. If the reviewer reports "⚠️ Cannot verify from diff" items, read references/cannot-verify.md before you mark the task complete.

Template: task-reviewer-prompt.md

4. The fix loop

The loop triggers when the review reports spec ❌, any Critical or Important finding, or a ⚠️ item you confirmed as a real gap.

When it triggers, read references/fix-loop.md and follow it. Read it again before every round, even if you read it earlier in this session: it holds the round cap, the breaker, and the ledger line each round writes, and your memory of it does not survive compaction.

5. Complete the task

When the review comes back clean — or every open finding is parked with a ruling at the cap — append the completion line to the ledger in the same message as your other bookkeeping:

  • Task <N>: complete (commits <base7>..<head7>, review clean)
  • Task <N>: complete (commits <base7>..<head7>, <K> parked) after a tripped breaker

Then mark the todo complete and move on. Never move to the next task while the review has open Critical/Important issues that are neither fixed nor parked-with-ruling at the cap.

Final Review

When every task is complete, read references/final-review.md and follow it. It covers the whole-branch review, its single fix wave, and how you finish.

Common Rationalizations

Excuse Reality
"Close enough on spec compliance" Reviewer found spec gaps = not done. Fix or hit the cap and adjudicate — those are the only exits.
"I'll fix it myself, dispatching is overhead" Controller fixes pollute your context and skip review. Resume the implementer.
"One more round will converge" Past the cap, rounds don't converge — the failure is structural. Adjudicate and route.
"The reviewer will just find something new anyway" Scoped re-reviews verify fixes; they cannot wander. New findings on untouched code go to the ledger, not the loop.
"This finding is obviously wrong, I'll drop it" You adjudicate only at the cap, and every ruling is a ledger entry. Silent discards are forbidden.
"The fix was small, skip the re-review" Unreviewed fixes are how regressions land. Every round ends with a scoped re-review.
"Reviews slow the loop down" The loop without reviews is just unverified churn. Reviews are the loop's brakes and steering.
"Ledger bookkeeping is overhead" The ledger is what survives compaction. Controllers without one have re-dispatched entire completed task sequences.
"The implementer spawned its own reviewer — free extra assurance" It's a duplicate seat reviewing the same diff; the task review is the gate. A worker-spawned reviewer is a defect to flag, not rigor.

Connect through AnyModel API

This skill runs on whatever model your agent uses. Point Claude Code, Codex or Cursor at AnyModel to run it on Claude, GPT, Gemini and more with one key — pay per token, no subscription.

More from the pack Superpowers

Brainstorming before you build

Finds out what you want to build and why: one question at a time, its understanding played back part by part, two or three approaches to pick from and a design doc, with no code until you approve it. Use for a new feature or change when the request leaves the essentials open.

Dispatching parallel agents

Splits independent problems between parallel subagents, each with one domain and a self-contained brief with scope, constraints and the report expected back, then checks their edits for conflicts and runs the full suite. Use when several unrelated things fail at once.

Executing plans

Runs a plan from writing-plans in the current session with no subagent per task: tasks in order through TDD, a ledger of progress and rulings that survives context compaction, and one fresh review of the whole branch at the end. Use when per-task review is not needed or there are no subagents.

Finishing a development branch

Finishes a branch: runs the tests, offers to merge locally, open a pull request or keep it as is, reruns the tests after a merge and removes its own worktree, and discards work only after you type "discard". Use when the implementation is done.

Receiving code review

Checks each review comment against the code before changing anything, clears up every unclear item first, pushes back with technical reasons and never performs agreement. Use when working through review feedback without blind fixes.

Requesting code review

Hands finished work to a fresh reviewer subagent through a template: the commit range, the plan or requirements, findings by severity and a verdict on merging. Use when a task or feature is done and a merge to main is next.

Systematic debugging

Keeps the agent from guessing at fixes: reproduce the bug and find the root cause first, test one hypothesis at a time, fix with a failing test, and question the design after three failed fixes. Use for any bug, test failure or unexpected behavior.

Test-driven development, Superpowers style

Drives work test-first: a test seen failing, then the least code that passes and a run of the whole suite; code written before its test is deleted. Use for features and bug fixes when the agent tends to add tests after the code.

Using git worktrees

Moves feature work into an isolated git worktree: checks whether the agent is already isolated, prefers the harness's own worktree tool, keeps the folder git-ignored, installs dependencies and runs a baseline test. Use when the work must not land on the branch you have open.

Verification before completion

Stops the agent from saying "done" or "tests pass" until it has rerun the check and read the whole output. Use whenever the agent reports success, commits or opens a pull request.

Writing implementation plans

Turns a spec into a plan for an implementer new to the codebase: files, signatures, interfaces between tasks, tests with exact values and TDD steps with each command's expected output. Use for a multi-step task when the spec is ready and no code exists yet.

Writing skills

Writes and edits skills the TDD way: a subagent first does the task without the skill and its mistakes are written down, then the smallest text that answers them goes in and the loopholes get closed. Use for a skill that other agents will load.